Opinion
No. 1851.
September 7, 1926.
Horace Van Everen, of Boston, Mass., for plaintiff.
William Quinby and Marcus B. May, both of Boston, Mass., for defendant.
In Equity. Suit by the Barber Colman Company against James Withnell. Decree dismissing bill.
This is a suit for the infringement of three patents, each to H.D. Colman, one, No. 977,166, dated November 29, 1910, which I shall refer to as the patent of 1910; another, No. 1,115,399, dated October 27, 1914, which I shall refer to as the patent of 1914; and a third, No. 1,442,776, dated January 16, 1923, which I shall refer to as the patent of 1923. The usual defenses are interposed, viz. noninfringement, and invalidity, if construed to cover defendant's machine, and also a suggestion that the 1923 patent is invalid, or limited, because of double patenting.
All three patents relate to preparing warps for weaving by mounting on them the necessary harnesses and reeds. This is done, either by tying each thread of the new warp to the corresponding thread of the old one, on which the reed and harnesses are still mounted, or by drawing each end of the new warp through its proper harness and reed. Machines for doing the first are called warp-tying machines; those for doing the second, warp-drawing machines. The defendant's machine neither ties nor draws the warp. It merely picks up the warp threads successively and presents them to an operative, who draws them in by hand. It is an aid to hand work, while the machines of the patents are fully automatic.
Mounting the harnesses and reeds on a warp has always been recognized as a necessary and troublesome piece of work. Until the advent of the machinery made by the plaintiff and its predecessors in business it was, on the evidence, done by hand. I remember as a boy seeing in the Fall River cotton mills lines of women and girls patiently and carefully drawing in, thread by thread, the new warps. In devising machines to perform this task, the plaintiff and those under whom it claims made a valuable contribution to the industry. If this result was brought about by the patents, and the claims thereof, here in suit, such patents and claims are entitled to friendly and sympathetic consideration and to a wide range of equivalents. The defendants do not admit that such is the fact, or that the claims here in suit are for the inventions to which that result is attributable. In complicated machines like these many factors enter into commercial success, and it is not easy, nor always possible, to say how much any given factor contributed.
A warp consists of many separate threads, or "ends," wound on a beam which is in effect a large spool. By passing a warp over a rounded bar and putting tension on it, all its threads can be brought into the same plane and for practical purposes into proper sequence; i.e., the threads are not crossed enough to interfere with successful weaving. But when this has been done the spaces between the threads are unequal, and this unevenness of spacing presented difficulties in designing a machine to draw or to tie warps. The first problem was to devise some mechanism for keeping the operating parts of the machine in contact with the leading warp thread. It often happens that, in stretching the warp over a bar, two threads will lie very close together, and, on the other hand, single threads will become widely separated (comparatively speaking) from their fellows. The machine must also be able to pick up but one of two closely adjacent threads, as well as a thread which was widely spaced.
All this had long been recognized at the time when the plaintiff's inventor, Colman, entered the field. As early as 1868 Biggs had taken out a British patent upon a machine for drawing in warps, in which he showed mechanism for handling one by one the warp threads. It worked upon a principle which is that of a successful machine, built and sold by the plaintiff's predecessor in business, the American Warp-Drawing Company, but which is basically different from that of the patents here in suit. There appear to have been two different solutions of the problem.
In this country several patents for warp-tying and warp-drawing machines, which long anticipated Colman, and which operate upon the same general principle as his machines, have been put in evidence. All of them either advanced the operating mechanism to the warp or — which amounts to the same thing — advanced the warp to the mechanism. They all had a thin, flat selector, with a smooth guiding surface, which engaged and glided on the leading warp thread and was provided with a projecting hook or notch, which caught the thread and carried it forward. This is shown in Rice (No. 176,194, April 18, 1876), in Sherman, Ingersoll Moore (No. 255,038, March 14, 1882), Sherman (No. 355,221, December 28, 1886), and Ingersoll (No. 590,008, September 14, 1897). In some of these — e.g., Sherman and Sherman, Ingersoll Moore — the selector was a thin circular disk having on its periphery a notch which caught the leading thread, and in both of these last patents there were means of adjustment to make the notch deep or shallow according to the size of the yarns being used. The point of these notches was wedge or chisel shaped; i.e., it had a sharp flat edge, adapted to enter between the leading thread and the thread next to it. If the notch were set too small, or the thread at the point where the selector met it happened to be oversize, which would often happen, it cannot be doubted that the edge of the notch caught in the thread and pulled it along. The same must have been true of the straight selector shown in the Ingersoll patent of 1897.
The first of the patents in suit, that of 1910, is upon a form of selector. In it the outer point of the notch is rounded into a needle-like point "sufficiently sharp and tapering to penetrate into the substance of a thread ner side of the point and seat itself within the which for any reason does not pass to the inthreads excepting those * * * small pocket." Patent, p. 1, line 53 et seq. "All enough to enter the hook opening are speared or impaled by the sharp point of the hook." Patent, p. 1, line 58 et seq. All the claims of this patent here in suit contain the word "impale" as describing the action of the hook point of this selector.
The defendant's selector consists of a straight feeler, into which is set a flat blade, the tip of which projects beyond the side of the feeler in such a way as to make a notch. It is exactly the same idea shown in the Sherman patent of 1886 and the Sherman, Ingersoll Moore patent of 1882, above referred to, except that in those patents the selector is circular, and in the defendant's it is straight. The notch point is chisel-shaped as those were, instead of needlelike as in the patent. Undoubtedly it catches in the thread when the thread is oversize or the notch is too small, just as the selectors did in the patents last referred to. But the defendant's device has not the "tapering," needlelike point for "spearing" or "impaling" a thread, which is the distinguishing characteristic of the plaintiff's patent. It differs from the patent very much as the patent differs from the prior art.
It follows that this patent is not infringed by the defendant, and as to it the bill should be dismissed.
The 1914 patent is for a machine for drawing warps. It has never been commercially manufactured and sold. Some of the inventions covered by this patent appear to have been made by Colman as long ago as October 24, 1894, when he filed an application for a United States patent, which by subsequent proceedings became merged into the application for this patent. At the expiration of this patent in 1931, Colman will have been claiming, and for practical purposes will have had, a monopoly of those inventions for 38 years.
The claims of this patent here in suit relate to a thread-controlled sensitive feed for the carriage. The twelfth may be regarded as the most typical. All the elements in this claim were old, and the combination was old, with the possible exception of the last element, which reads as follows: "And a sensitive feeler finger arranged to bear against the foremost thread and to control the moving means." The feeler finger is a long and delicately balanced lever, under tension of a light spring, which tilts the top forward. The top contacts with the leading warp thread. The advance of the warp carriage on which the feeler is mounted overcomes the spring and swings back the top of the feeler by the pressure of the warp thread. This motion of the feeler throws out of action a pawl, by which (acting on a ratchet) the carriage is advanced, thereby bringing the carriage to rest. When the leading thread has been taken away, the feeler, under the influence of the spring, retakes its advanced position. This frees the pawl to operate again, and the carriage moves forward until arrested as above described. The feeler mechanism is entirely independent of the thread-selecting mechanism, which may be disregarded in this discussion.
In the defendant's machine the functions of the feeler and the selector are combined into the feeler selector above described. It is mounted on a delicately balanced lever having a light spring, which tends to draw the feeler part forward, as is the case in the Colman patent. As the carriage advances, the forward motion of the selector feeler becomes limited at a certain point in its cycle by contact with the warp, and it thereby throws out of operation the machinery (which consists of pawl and ratchet) by which the carriage is moved. To some extent the two mechanisms involve the same underlying principle, viz. a pivoted feeler held back or limited in motion by the leading warp thread as the carriage advances, until it finally throws the feed out of action. Beyond this they are radically different. The defendant's machine is infinitely simpler and more ingenious, displaying invention of a high order as compared with the complicated and commercially unsuccessful machine of the patent.
The first question under this patent is, I think, whether mechanism for controlling the movement of the carriage by the contact of such a pivoted feeler with the leading warp thread was new with Colman. If so, he would be entitled to such a wide range of equivalents and modifications as would include the defendant's device. That this is the fact the defendant disputes. The references principally relied upon are the patents to Biggs, to Sherman, Ingersoll and Moore, and to Shackleton and Binns, above referred to. In the Biggs patent the carriage has a number of revolving cylinders with screw threads on their surfaces, which are set at a slight angle to the plane of the warp and progressively engage the warp threads as the carriage advances. The rotation of these cylindrical screws separates and spaces the warp threads. Finally, as the carriage continues to advance, the leading warp thread is dropped off the end of the last cylinder, and rests upon a pivoted finger (called "trigger" in the patent), which it depresses. This operates to arrest the motion of the carriage, which remains stationary until the thread upon the "trigger" has been removed. The removal of that thread allows the "trigger" to retake its normal position, and the forward movement of the carriage is resumed, until another thread drops onto it. Biggs patent, p. 2, line 35 et seq. This patent discloses the idea of controlling the movement of the carriage by the pressure of the leading warp thread on a pivoted lever, although the way in which the idea is worked out is very different from that embodied in the plaintiff's machine, because it is not the advance of the carriage which causes the pressure on the feeler, but the tension of the thread. In the Biggs machine the carriage is not moved by a ratchet and pawl driven by the power on the machine, but is drawn forward by a weight controlled by a gear which carries a check stud, which is engaged by the feeler. In the plaintiff's brief the Biggs patent is distinguished on the ground, inter alia, that the advance of its carriage is "fixed and regular"; the implication being that in the 1914 patent the advance is not of that character. No such limitation is found in the claims. In all machines the advance is step by step, the length of the step depending, in the plaintiff's machine, on the size of the ratchet teeth and the movement of the pawl, and in the Biggs on the size and arrangement of the controlled gear.
In the Sherman, Ingersoll and Moore patent, the leading thread, after being seized and drawn out of line by the selector, pulls a lightly balanced arm, which it rotates in such a way as to throw out of action the stopping mechanism, thereby allowing the machine to continue in operation. If the selector fails to catch and bring forward a warp thread, the arm is not rotated, and the stop motion comes into action and stops the machine. Here again we see the action of the machine continued or arrested with reference to the leading warp thread, but not by pressure caused by the advance of the carriage. The Shackleton Binns patent also shows mechanism governed by the leading warp thread for stopping the entire machine, and claim 5 of this last patent is a combination of the basic elements "and devices to operate the catches to stop the machine when a warp thread breaks or is not caught by the needle."
Summarizing the prior art, it shows that the idea of controlling the feed through mechanism depending for its operation upon the leading warp thread was not new at the time of the Colman application, and had been worked out in several machines.
The next question is whether the defendant's machine infringes the patent, construed in the light of the prior art, and this requires consideration of the dissimilarities between the two devices. The Colman feeler was, as I have said, a long delicately hung lever, which was intended to be in constant contact with the leading warp thread while the machine was in use. It yielded to the pressure of the warp when the carriage advanced, and thereby threw the feed mechanism out of action. It had no motion except a slight backward and forward one around its pivot, and it performed no function, except that of feeler. The defendant's feeler selector is carried on a short lever mounted on an arm which is moved by a revolving cam. It moves lengthwise as well as ocillates. At one part of the cycle its tail is drawn over a stationary cam giving a circular movement to the selector. A large part of the time it is not in contact with the warp at all, being sometimes entirely to one side of the warp, and at other times (when it is carrying forward a thread which it has seized) some distance ahead of the warp. It has a sliding motion past the warp. At one stage in its movement, it responds to the action of a light spring and is pulled against the leading warp thread, which momentarily limits its motion in that direction. If the carriage be close to the thread, so that the pull of the spring is resisted by it, and the feeler is thereby kept nearly vertical to the warp, the feed mechanism of the carriage is kept out of action. If, on the other hand, the carriage is so far from the warp that the feeler is on a decided slant when it comes into contact with the leading warp thread, the feed mechanism of the carriage operates, and the carriage moves up, until the feeler becomes so nearly vertical when it touches the warp that it again throws the feed out of action. After contacting with the warp, the feeler has the sliding motion past it above referred to, in the course of which it seizes by its notch above described the leading warp thread. Having done so, it separates that thread and pulls it away from the others, and presents it to be acted upon by the other parts of the machine.
Whether the defendant has infringed depends primarily upon whether he has used the mechanical idea of this Colman patent, or, as stated in the plaintiff's brief, "no court would hold a defendant as an infringer if he had not appropriated the invention, even though the language of the claim, as mere language, might apply to his machine" (page 64). There is, I think, a radical difference between a feeler in constant contact with the warp, swinging forward and back around a stationary pivot, and one which moves in every direction and during most of its cycle is entirely out of contact with the warp and makes use of the warp only to limit its forward play at a certain point in its cycle. It had been shown in the prior art that the leading warp thread was the thing by which the feed of the carriage must be controlled, and that it could be used by its pressure on a pivoted lever to accomplish that result. Colman showed that the direct pressure of the feeler against the warp caused by the forward motion of the carriage could be used to control the feed. In the defendant's device the pressure of the feeler against the warp is set up, not by the movement of the carriage, but by the rotation of the cam, which projects the feeler selector into the line of the warp, whereupon the spring above referred to pulls it against the leading warp thread. While superficially the two devices appear to involve the same principle of operation, they are in reality basically different. In other words, the defendant's machine is, in my opinion, not an improvement or development of the Colman idea, but an independent solution of the problem. I doubt whether any idea disclosed for the first time in the Colman patent of 1914 is used in the defendant's feeler selector, and I am clear that to give that patent an interpretation broad enough to cover the defendant's device would extend it far beyond its proper scope. The merits of this patent have been exaggerated, I think, by the plaintiff. Certain of its claims, if broadly construed, can be read on the defendant's machine; but, if given the same breadth, they read on machines of the prior art. The plaintiff cannot have it both ways — narrow for the prior art and broad for the defendant.
Holding these views, it is unnecessary for me to discuss the precise language of the various claims in suit. I am of opinion that none of them are infringed by the defendant's machine. Nor is it necessary to pass upon the defendant's contention that the claims are too vague and indefinite to be valid.
The last patent in suit is the Colman patent of 1923, and of this patent only one claim, the 218th, is in issue. This patent was in the Office about nineteen years, and during that interval the patent to Moore (No. 1,259,546, dated March 19, 1918), under which the defendant's machine is built, was issued. If claim 218 be construed as covering all stop motions controlled by the warp thread, it unquestionably lacked novelty, and is for that reason invalid. To avoid this result the plaintiff construes it as covering only mechanism which, while stopping the travel of the carriage, permits the rest of the machine to continue. While, in view of the patentee's use elsewhere of the word "suspend" in connection with a stop motion which stopped the entire machine, and also because of the improbability that such an invention would be referred to in only one of the 293 claims which the patent includes, the point is not free from doubt, I have reached the conclusion that the claim ought to be interpreted in accordance with the plaintiff's view. The defendant still contends that the claim, so construed, is invalid, because anticipated, and because it is "nebulous" and "indefinite," and for a result or idea, rather than for specific mechanism.
The first four elements of this claim in combination are shown in Biggs too clearly to require discussion. The only room for doubt is whether the other element is there also. The carriage in that patent is drawn forward by a weight; its propulsion is entirely independent of the power driving the machine. Its feed is regulated solely by the escapement gear controlled by the "trigger" or feeler. As I understand the patent, the screwlike selectors continue to revolve, whether the carriage is moving or at rest. If this be so, the claim in suit is precisely readable on Biggs.
It is argued by the plaintiff that the Biggs patent is for an inoperative machine, that it is a mere paper patent, and a foreign one at that, and that it cannot be availed of to invalidate a later United States patent. I quite agree that a patent for a meritorious invention ought not to be defeated by an inoperative and practically worthless patent of prior date. The law to that effect is well and justly settled. But when both patents are upon complicated machines, involving many different inventions, and requiring great practical knowledge and skill to adapt them to the market, lack of commercial use or success is not strong evidence that the inventions disclosed were not genuine and meritorious. For instance, the plaintiff's 1914 patent has 107 claims. The machine therein described was not commercially successful; but that fact has not been, and I do not think could be, urged with much force against the validity of the patent. The general organization of the plaintiff's machine is similar to that of Biggs'; i.e., a frame for holding the warp, and a carriage carrying the operating mechanism and advancing on guides towards the warp, regulated in its movement by the leading warp thread. These ideas, which appear to have originated with Biggs, have been adopted in many subsequent machines for the same purpose. As has been already noticed, certain basic ideas of Biggs were used commercially by the plaintiff's predecessor. It would be going much too far, I think, to say that the Biggs patent should be disregarded in considering the prior art, and Colman allowed to patent a combination plainly shown in this very art more than 50 years before.
There is also the question which I recently considered in a case involving a very different art (Johnson Automatic Scale Co. v. Ginn, Dist. Ct. Mass. opinion Jan. 29, 1926, p. 8), that restricting the action of a stop motion to a particular part of a machine did not, except where the problem presented special difficulties, constitute invention, but was a matter of machine design. The language of Judge Brown in Robinson v. Tubular Woven Fabric Co. (D.C.) 248 F. 526, at page 541, affirmed (C.C.A. 1st) 254 F. 304, 166 C.C.A. 44), is very appropriate here. This claim comes after the 1914 patent in suit, and is still further limited by it. In my opinion, claim 218 is invalid.
Decree dismissing bill, with costs.